LangChain, a framework for building large language model (LLM) applications, has released version 0.2 with several improvements and new features. Here’s a quick rundown:
- Enhanced Stability and Security: The langchain package is now decoupled from langchain-community for better stability and security.
- Improved Documentation: Versioned docs with better discoverability allow you to easily find information relevant to your specific LangChain version.
- LangGraph Takes Center Stage: LangGraph is now the recommended way to build agents, simplifying cycle and memory management while offering easy customization.
- Standard Chat Model Features: Standardized tool calling support and a structured output interface enhance chat model functionality.
- Asynchronous Processing and Streaming: Async support for core abstractions and a new Event Streaming API enable smoother workflows.
- Rich Partner Ecosystem: Over 20 partner packages across Python and JavaScript offer extended functionalities.
LangSmith Gains GDPR Compliance, RBAC, and Pairwise Evaluation
LangSmith, the unified developer platform for building, testing, and monitoring LLM applications, has received several upgrades:
- GDPR Compliance: LangSmith now adheres to GDPR regulations, ensuring data privacy for users and enterprises.
- Role-Based Access Control (RBAC): Enterprise users can assign roles with granular permissions to control resource access within their organization.
- Enhanced API Keys: Personal Access Tokens and Service Keys provide improved access control for users and services.
- Pairwise Evaluation: This feature helps choose the best output from LLM generation tasks where a single “correct” answer might not exist.
- Improved Prompt Management: Personal prompts are now separate from LangChain Hub, allowing for easier organization and version tracking.
Speak the Lang: Real-World Use Cases of LangChain, LangSmith, and LangGraph
This section highlights successful applications built with LangChain’s suite of tools:
- Multi-agent Research Assistant:A step-by-step guide demonstrates how LangGraph and GPT Researcher can collaborate to create an autonomous research assistant, generating multi-page reports on various topics.
- LangGraph Customer Support Bot: This example showcases the creation of a customer support bot that can research and manage bookings using LangGraph.
- Multi-agent RAG for Technical Blogs: The AI Makerspace team demonstrates how LangChain and LangGraph can be used to coordinate agents for building technical blogs.
LangChain Evaluations: Exploring GPT-4o and RAG
LangChain provides tools to evaluate and optimize LLM applications:
- Testing GPT-4o with LangSmith: This video explores the performance of OpenAI’s new GPT-4o model compared to older versions using a simple RAG application.
- Evaluating Intermediate Steps in RAG Pipelines: Learn how to isolate and evaluate the outputs of each step in your RAG pipeline, enabling better debugging and performance optimization.
LangChain Partners and Community
LangChain fosters a rich ecosystem with partners and a thriving community:
- New Partner Packages: LangChain offers integrations with Hugging Face and Qdrant through partner packages.
- Multimodal RAG with Redis: This blog post explores the benefits of multimodal RAG, allowing models to process and reason across text and images.
- LangChain Wins MongoDB’s AI App Framework Partner of the Year Award
From the Community
Videos:
- Building LLM Agents with Tool Use by Jay Alammar at Cohere
- Flow engineering with LangGraph: GPT Newspaper by Eden Marco (with repo built by Rotem Weiss)
Blogs:
- Two underestimated LangChain features to create production-ready configurable chains by Rav on Metadocs
- Using Server-Sent Events (SSE) to stream LLM responses in Next.js by Rishi Jain, founder at LaunchFa.st
- LangChain Chatbot Framework With Retrievers by Cobus Greyling, Chief Evangelist at Kore AI
- Tips for Building a RAG Pipeline with NVIDIA AI LangChain AI Endpoints by Amit Bleiweiss, Sr. Data Scientist at NVIDIA
Courses:
- Prompt engineering LinkedIn course by Harpreet Sahota
GitHub Projects & Notebooks:
- Build a RAG system with Llama 3B-Instruct for your PDFs by Maria Khalusova at Unstructured
- Cohere toolkit for building RAG apps
- WebRTC AI Voice Chat with LangChain
Learn More about LangChain
For detailed information on these updates and to explore LangChain’s functionalities, refer to the official resources.